METHOD AND SYSTEM FOR VOICE RECOGNITION MENU NAVIGATION

• US 20060200348A1
• Filed: 05/22/2006
• Published: 09/07/2006
• Est. Priority Date: 04/08/2002
• Status: Active Grant

• Current Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Original Assignee: AT&T Labs Incorporated (AT&T, Inc.)
• Inventors: Knott, Benjamin Anthony, Martin, John Mills, Bushey, Robert Randal, Smart, Tracy Leigh
• Granted Patent: US 7,502,742 B2
• Time in Patent Office: Days
• US Class Current: 704/246
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue

Abstract

A method and system for error prevention and recovery of voice activated navigation through a menu having plural nodes provides situation dependent utterance verification by relating confirmation to utterance determination confidence levels. In one embodiment, a high confidence level results in implicit confirmation, a medium confidence level results in explicit confirmation and a low confidence level results in a concise interrogative prompt of a single word that requests the user to repeat the utterance. In situations where voice recognition is difficult, dual modality with DTMF navigation is provided as an option for menu selections.

Claims

1-27. -27. (canceled)

28. A method for facilitating navigation of a voice recognition menu, the method comprising:

in a voice recognition service having plural nodes, receiving from a user a user utterance associated with a particular one of the plural nodes;

electronically attempting to recognize the user utterance;

electronically determining a statistical likelihood that the attempt to recognize the user utterance is correct; and

evaluating the determined statistical likelihood with respect to one or more predetermined confidence levels.

29. A method according to claim 28, wherein evaluating the determined statistical likelihood with respect to one or more predetermined confidence levels comprises electronically classifying the user utterance into one of a plurality of confidence levels based at least on the determined statistical likelihood and a set of quantitative confidence level ranges corresponding with the particular node, wherein different ones of the plural nodes have different confidence level ranges.

30. A method according to claim 29, wherein classifying the user utterance into one of a plurality of confidence levels comprises classifying the user utterance as a high confidence utterance, a medium confidence utterance, or a low confidence utterance.

31. A method according to claim 30, further comprising:

if the user utterance is classified as a high confidence utterance, advancing the user to a menu node determined to be associated with the high confidence utterance;

if the user utterance is classified as a medium confidence utterance, prompting the user for confirmation of a menu node determined to be associated with the medium confidence utterance; and

if the user utterance is classified as a low confidence utterance, requesting the user to restate the user utterance.

32. A method according to claim 31, further comprising obtaining implicit confirmation of the high confidence utterance before advancing to the associated menu node.

33. A method according to claim 32, wherein obtaining implicit confirmation further comprises:

stating to the user the menu node determined to be associated with the high confidence utterance;

advancing to the determined menu node if the user either fails to respond or confirms the menu node; and

requesting the user to restate the user utterance if the user refutes the determined menu node.

34. A method according to claim 31, wherein prompting the user for confirmation of a menu node determined to be associated with the medium confidence utterance comprises:

stating to the user the menu node determined to be associated with the medium confidence utterance;

advancing to the determined menu node if the user responds with affirmation; and

requesting the user to restate the user utterance if the user refutes the determined menu node.

35. A method according to claim 28, wherein attempting to recognize the user utterance comprises attempting to identify a node selection intended by the user.

36. A method according to claim 28, further comprising:

selecting a prompt to communicate to the user based at least on the classification of the user utterance; and

communicating the selected prompt to the user.

37. A method according to claim 28, further comprising:

if the user utterance is classified into a particular confidence level, requesting the user to restate the user utterance and attempting to recognize the restated user utterance;

counting the number of attempts to recognize the user utterance; and

automatically directing the user to a DTMF menu after a predetermined number of attempts.

38. A system for facilitating navigation of a voice recognition menu having plural nodes, the system comprising:

a computerized voice recognition response unit interfaced with a telephone network, the voice recognition response unit operable to receive from a user a user utterance associated with a particular one of the plural nodes, and attempt to recognize the user utterance;

a computerized confidence engine interfaced with the voice recognition response unit and operable to;

determine a statistical likelihood that the attempt to recognize the user utterance is correct; and

classify the user utterance into one of a plurality of confidence levels based at least on the determined statistical likelihood and a set of quantitative confidence level ranges corresponding with the particular node, wherein different ones of the plural nodes have different confidence level ranges.

39. A system according to claim 38, further comprising an interrogate engine interfaced with the voice recognition response unit and operable to command an implicit confirmation for a first confidence level, an explicit confirmation for a second confidence level and an interrogative prompt for a third confidence level.

40. A system according to claim 38, wherein the confidence engine is operable to classify the user utterance as a high confidence utterance, a medium confidence utterance, or a low confidence utterance.

41. A system according to claim 40, further comprising an interrogate engine operable to cooperate with the confidence engine to:

if the user utterance is classified as a high confidence utterance, advance the user to a menu node determined to be associated with the high confidence utterance;

if the user utterance is classified as a medium confidence utterance, prompt the user for confirmation of a menu node determined to be associated with the medium confidence utterance; and

if the user utterance is classified as a low confidence utterance, request the user to restate the user utterance.

42. A system according to claim 41, where the confidence engine obtains implicit confirmation of the high confidence utterance before advancing to the associated menu node.

43. A system according to claim 42, wherein obtaining implicit confirmation further comprises:

stating to the user the menu node determined to be associated with the high confidence utterance;

advancing to the determined menu node if the user either fails to respond or confirms the menu node; and

requesting the user to restate the user utterance if the user refutes the determined menu node.

44. A system according to claim 41, wherein prompting the user for confirmation of a menu node determined to be associated with the medium confidence utterance comprises:

stating to the user the menu node determined to be associated with the medium confidence utterance;

advancing to the determined menu node if the user responds with affirmation; and

requesting the user to restate the user utterance if the user refutes the determined menu node.

45. A system according to claim 38, wherein attempting to recognize the user utterance comprises attempting to identify a node selection intended by the user.

46. A system according to claim 38, further comprising an interrogate engine interfaced with the voice recognition response unit and operable to:

select a prompt to communicate to the user based at least on the classification of the user utterance; and

communicate the selected prompt to the user.

47. A system according to claim 38, further comprising an interrogate engine operable to cooperate with the confidence engine to:

if the user utterance is classified into a particular confidence level, prompt the user to restate the user utterance and attempting to recognize the restated user utterance;

count the number of attempts to recognize the user utterance; and

automatically direct the user to a DTMF menu after a predetermined number of attempts.

48. A system comprising at least one computer-readable medium including computer-executable instructions for facilitating navigation of a voice recognition menu having plural nodes, the computer-executable instructions comprising:

instructions for receiving from a user a user utterance associated with a particular one of the plural nodes;

instructions for attempting to recognize the user utterance;

instructions for determining a statistical likelihood that the attempt to recognize the user utterance is correct; and

instructions for classifying the user utterance into one of a plurality of confidence levels based at least on the determined statistical likelihood and a set of quantitative confidence level ranges corresponding with the particular node, wherein different ones of the plural nodes have different confidence level ranges.

49. A method for facilitating navigation of a voice recognition menu, the method comprising:

in a voice recognition service having plural nodes, receiving a user utterance associated with a particular one of the plural nodes;

electronically attempting to recognize the user utterance;

electronically determining a statistical likelihood that the attempt to recognize the user utterance is correct; and

electronically classifying the user utterance into one of a plurality of confidence levels, including determining a confidence level corresponding to the determined statistical likelihood, wherein confidence levels corresponding to various statistical likelihoods differs for different ones of the plural nodes such that the determination of the confidence level corresponding to the determined statistical likelihood depends on the particular node.

50. A method according to claim 49, wherein classifying the user utterance into one of a plurality of confidence levels comprises classifying the user utterance as a high confidence utterance, a medium confidence utterance, or a low confidence utterance.

51. A method for facilitating navigation of a voice recognition menu, the method comprising:

in a voice recognition service having plural nodes, receiving a user utterance associated with a particular one of the plural nodes;

electronically attempting to recognize the user utterance;

determining a statistical likelihood that the attempt to recognize the user utterance is correct;

if the determined statistical likelihood is below a particular threshold, requesting the user to restate the user utterance and attempting to recognize the restated user utterance;

counting the number of attempts to recognize the user utterance; and

automatically directing the user to a DTMF menu after a predetermined number of attempts.